At Mon, 06 Jan 2014 17:33:47 +0900, Takashi Sakamoto wrote:
Hi Iwai-san,
1.snd-firewire-speakers and snd-oxfw
I don't see why these make things difficult to merge? They sound rather like the functionalities don't conflict.
I think it better not to touch existed drivers unless they have some bugs.
Actually I can merge two drivers with some selection statements, many changes for card structure and for PCM functionality, many codes for MIDI and hwdep functionaloty. But for snd-firewire-speakers these modifications are needless and not due to bugs.
...On the other hand, I also think it good to merge these two drivers because they support the same chipset. But this idea have a risk to put some new bugs into snd-firewire-speakers.
Yes, having two different drivers for the same hardware leads nothing but confusion, judging from the past experience.
Note that since I haven't seen any ack from Clemens, I didn't take previous patches yet.
Yes. I've posted my patches with RFC, without merge-request. I describe the reason.
OK.
Devices which snd-fireworks, snd-bebob and snd-oxfw supports are categorized in 'GenericAVC' by FFADO. This means the devices are controlled by the same way, AV/C commands. At first, I considered to develop 'snd-genericavc', which perform like OS X implementation. But there were some anxieties for device quirks.
In this reason, I decide to develop them in the same time. As a result, I confirm each chipset/firmware has it own quirk. So now I have a confidence that it's difficult to develop 'snd-genericavc'.
And it's a bit too late for 3.14, we have plenty of time for review
and discussion for 3.15.
Currently I have no will to merge my patches because I need more discussions between ALSA/FFADO developers.
My patches add ALSA to handle devices which FFADO supports for its streaming functionality. It may bring a bit complexity to users. I need to adjust two projects in advance of merge-request.
Fair enough.
thanks,
Takashi